Cell Motility and the Cytoskeleton 18:12%130 (1991)
Calcium Sensors in Sea Urchin Sperm Flagella Charles J. Brokaw Division of Biology, California Institute of Technology, Pasadena
The asymmetry of ATP-reactivated flagellar bending waves of Triton-demembrated sea urchin spermatozoa has been measured over a range of free Ca2+ ion concentrations from to lop4 M. Detailed examination of the gradual response of asymmetry to Ca2+ ion concentration over this wide range indicates the presence of two Ca2+ sensors. A high-affinity sensor operates at Ca2+ concentrations near M. A lower-affinity sensor operates at Ca2+ concentrations above M, in the typical range for calmodulin-mediated responses. Incubation of demembranated sperm flagella at high Ca2+ concentrations to release calmodulin is required to enable these Ca2+ responses to be observed. This treatment also causes a decrease in the apparent affinity of the flagella for calmodulin, as determined by measuring the increase in asymmetry in response to addition of exogenous calmodulin at low Ca2+ Concentration. Key words: calmodulin, motility, spermatozoa
INTRODUCTION The ATP-reactivated movement of Triton-demembranated sea urchin sperm flagella can be regulated by the Ca2+ ion concentration of the reactivation solutions, with higher Ca2+ ion concentrations causing increased asymmetry of the flagellar bending waves, without any other major changes in bending wave parameters [Brokaw et al., 1974; Brokaw, 19791. Although a function for this Ca2+ sensitivity has not been definitively established, its involvement in chemotactic response of spermatozoa to egg jelly peptides has been suggested [Ward et al., 19851. The response to Ca2+ ion concentrations is difficult to study with demembranated sea urchin sperm preparations that are prepared exclusively using solutions containing EGTA to maintain very low Ca2+ ion concentrations (“potentially asymmetric” spermatozoa [Gibbons and Gibbons, 19801). Even when reactivated at very low Ca2+ concentrations, such sperm preparations display a high level of bending wave asymmetry, which increases to levels that are too extreme to measure accurately as the Ca2+ concentration is increased [Brokaw, 19791. Exposure to millimolar Ca2+ ion concentrations during or after demembranation can produce “potentially symmetric” sperm flagella that generate nearly 0 1991 Wiley-Liss, Inc.
symmetric flagellar bending waves at low Ca2+ ion concentrations and demonstrate gradually increasing asymmetry as the Ca2+ ion concentrations in the reactivation solutions are increased [Brokaw et al., 19741. This effect of millimolar Ca2 concentrations is time and temperature dependent, requires the presence of detergent [Okuno and Brokaw, 19801, results in release of calmodulin from the sperm flagella, and sensitizes the flagella to exogenous calmodulin [Brokaw and Nagayama, 19851. The exploration of the calmodulin and calcium sensitivity of flagella by Brokaw and Nagayama [ 19851 utilized spermatozoa of Lytechinus pictus. Spermatozoa from this sea urchin species show, in addition to increased asymmetry, an attenuation or damping of bending wave amplitude in the distal region of the flagellum as Ca2+ ion concentration is increased [Okuno and Brokaw, 19801. This effect interferes with measurement of bending wave asymmetry at higher Ca2+ concentrations. Spermatozoa from another sea urchin, Strongylocentrotus purpurutus, show much less increase in bend+
Received July 20, 1990; accepted October 25, 1990. Address reprint requests to Dr. C.J. Brokaw, Division of Biology 156-29, Caltech, Pasadena, CA 91 125.
124
Brokaw
ing wave attenuation at high Ca2+ ion concentrations. However, initial attempts to demonstrate calmodulinsensitivity of S. purpurutus spermatozoa following extraction with millimolar Ca2 solutions indicated that neither the procedures used for L . pictus spermatozoa [Brokaw and Nagayama, 19851 nor the procedures used earlier for demembranation and reactivation of S. purpurutus spermatozoa [e.g., Brokaw, 19861 gave good results. Development of a more optimal procedure for preparation of “potentially symmetric” sperm flagella with good calmodulin sensitivity from S. purpurutus has now made possible a more detailed examination of the relationship between flagellar asymmetry and concentrations of Ca2+ ions and calmodulin in this species. In particular, the ability of these flagella to respond gradually to changes in Ca2+ ion concentration over a range of more than 4 decades of Ca2+ ion concentration [Brokaw et al., 19741 has been examined more fully, since this behavior does not resemble that expected for a simple Ca-binding receptor. This behavior has been clarified by making measurements at intervals of 0.5 pCa unit, rather than just at intervals of 1 pCa unit, and by using a new calcium buffer, BAPTA, to confirm results obtained with solutions buffered with EGTA and EDTA. This increased data collection has been made possible by the evolution of efficient computerized methods for analysis of photographic records of sperm flagellar bending. +
MATERIALS AND METHODS
The methods used here for preparation and reactivation of spermatozoa from the sea urchin, S. purpurutus, differ from earlier methods used in this laboratory [Brokaw, 19861 but have been partially described in work reported by Tombes et al. [1987] and by Omoto and Brokaw [1989]. The sea urchins were induced to shed gametes by injection of 0.6 M KC1, and “dry” spermatozoa were collected from the aboral surface of the urchin, diluted with an equal volume of cold 0.5 M NaCl, and stored on ice until used. Demembranation was performed by mixing 1 p1 of the stock sperm suspension with 50 pl of demembranation solution containing 0.15 M KC1, 10 mM Tris buffer, 1 mM dithiothreitol (DTT), 2 mM MgSO,, 0.2 mM EGTA, 0.2 mM ATP, 10 pM CAMP, and 0.04% Triton X-100. This solution, and all of the other solutions used in this work, were adjusted to pH 8.2. In contrast to spermatozoa of L. pictus, spermatozoa from S. purpuru~ K consistently S show good reactivation without exposure to solutions containing CAMP, and it has not been possible to show any consistent benefits of including CAMP and ATP in the demembranation solutions unless the pH of the demembranation solution is reduced to about 7.0. These components were included routinely in the de-
membranation solution with the hope that this would increase the uniformity of the results obtained with different sperm samples. After 30 seconds incubation in demembranation solution, 450 pl of Ca-extraction solution containing 0.15 M KCl, 10 mM Tris buffer, 1 mM DTT, 2.2 mM CaCl,, and 0.01% Triton X-100 was added, and the mixture was incubated for an additional 30 sec. Ten microliters of this extracted sperm suspension was then diluted into 1.0 ml of reactivation solution for observations. All reactivation solutions contained 0.25 M potassium acetate, 10 or 20 mM Tris buffer, 1 mM DTT, and 0.5% polyethyleneglycol. ATP, MgSO,, CaCl,, EGTA, EDTA, and BAPTA were added as needed for particular experiments. Most of the experiments reported here used reactivation solutions containing 0.1 mM MgATP2+ and 0.3 mM Mg2+. Recipes for these solutions, containing desired Ca2+ ion concentrations, were calculated as described by Brokaw [1986], using the constants given there for EDTA and EGTA. The constants given by Tsien [1980] for BAPTA were used or modified for the higher ionic strength of the reactivation solutions used here, leading to a value of 6.5 for the pK,, for BAPTA [Harrison and Bers, 19871. These calculations assumed a level of calcium contamination in the solutions of 20 pM. EGTA, EDTA, and bovine brain calmodulin were purchased from Sigma and BAPTA was from Calbiochem. A drop of demembranated sperm suspension in reactivation solution was placed on a microscope slide and illuminated with darkfield strobe illumination, using a temperature-controlled stage at 18°C in a room maintained at 18°C. Photographs were taken at a magnification of 80 X and a flash rate of 120 or 150 flashes per second, on film moving at 0.25 m/sec in a 35 mm oscilloscope camera [Brokaw, 19861. A series of images from each photograph was digitized, either manually as in earlier work or with a new computerized digitization procedure [Brokaw, 19901, to obtain data sets that were analyzed by computerized methods [Brokaw, 19841 to obtain parameters for the flagellar bending waves, as in earlier work. RESULTS Parameters of Flagellar Motility of S. purpuratus Spermatozoa
Table I presents unpublished data from the experiments of Tombes et al. [ 19871 on live spermatozoa of S. purpurutus. Seven sperm samples were measured under control conditions, in artificial sea water at 18°C. This table also presents data for seven samples of demembranated spermatozoa, reactivated at an MgATP concentration of 0.5 mM and intermediate Ca2 concentration +
Calcium Sensors in Flagella
125
TABLE I. Parameters of Sperm Flagellar Motility Demembranated and 0.5 mM MgATP-reactivated spermatozoa (7 samples; 417 cells)
Live spermatozoa (7 samples; 153 cells) Mean of sample means
Std. error of mean
Mean sample std. dev.
Mean of samDle means
Std. error of mean
Mean sample std. dev.
38.4 1.84 180 28.7 0.35 1.18
0.5 0.05 6 1.3 0.03 0.06
2.7 0.11 15 2.0 0.17 0.21
33.6 2.15 185 28.1 0.27 1.33
0.3 0.06 5 0.6 0.05 0.11
1.7 0.19 16 1.8 0.19 0.28
Frequency (s-I) Bend angle (rad) Sliding velocity (rad s - ' ) ~ Wavelength (pm) Attenuation' Asymmetry (rad)'
"Velocity of metachronal sliding in propagating bends; average magnitude for principal and reverse bends. 'Attenuation measures decrease in curvature of bends as they propagate along flagellum [Tombes et al., 19871. 'Asymmetry calculated as difference between absolute values of principal and reverse bend angles [Brokaw, 19791.
(pCa 6.5). Under these conditions of demembranation and reactivation, the parameters of reactivated motility are very similar to those of intact spermatozoa, except for a somewhat lower frequency, which is combined with an increased bend angle so that there is no significant difference in sliding velocity. In the previous study [Brokaw, 19791, data were presented showing that Ca2 concentration had very little effect on parameters of flagellar motility, other than the asymmetry of the bending waves. Those experiments were performed at 0.1 mM MgATP, at relatively low beat frequencies (ca. 10 Hz) because the reactivation solutions contained methyl cellulose and KCl rather than K acetate. In the experiments reported here, at 0.1 mM MgATP with 0.3 mM Mg2+, the beat frequencies were approximately 19 Hz. As shown in Figure 1, under these conditions also, there is very little variation in sliding velocity, wavelength, or bending wave attenuation over the range from pCa 9.25 to 4.75. There is a slight decrease (ca. 10%) in beat frequency at higher Ca2+ concentrations (not shown), but this is compensated by an increase in bend angle, so that the sliding velocity remains essentially constant. However, at the higher MgATP concentration of 0.5 mM, which gives beat frequencies and sliding velocities more comparable to those of live spermatozoa, there are significant changes in some of these other bending wave parameters with pCa, especially at pCa values less than 7 (Fig. 1). To avoid complications resulting from these changes, effects of Ca2+ concentration on asymmetry were evaluated at 0.1 mM MgATP as well as at 0.5 mM MgATP.
0
8
+
The Effect of Ca2+ Ion Concentration on Flagellar Asymmetry
Figure summarizes the effect Of tion on the asymmetry of flagellar bending of demembranated spermatozoa of S . purpuratus reacti-
I 20 I
5 .c1
2 5 2
0 0
I
I
I
I
I
r
0
0.8 0.6
0
:
0
0.4
0
c.
-
0.2 0.0
1 I
250 r
0 Q)
a*
200-
2 2 150-
v
_0 0
5
>
100-
I
I
I
I
1
I
9
8
7
6
5
4
PCa Fig. 1. Effects of Ca'+ concentration on some bending wave parameters of ATP-reactivated spermatozoa from S. purpuratus. Solid points are from experiments at 0.1 mM MgATP; open points are from experiments at 0.5 mM MgATP. Each point represents a mean for measurements on approximately 40 spermatozoa. Values of bending wave asymmetry from these experiments are included in Figures 2 and 3.
126
Brokaw 0 0
S 0
O
0 0
0 0
0 ”
01 9
8
5
6
7
0
0
4
PCa
I
10
Fig. 2. Mean values of flagellar bending wave asymmetry from experiments with 7 different samples of S. purpurutus spermatozoa, reactivated with 0.5 mM MgATP and 1 or 2 mM M g 2 + . Results from experiments with 4 sperm samples, with Ca2 concentrations buffered with EGTA at pCa 7 and higher, and with EDTA at pCa less than 7 are shown by open squares. Results from experiments with another 3 sperm samples are shown by circles, with Ca2+ concentrations buffered with BAPTA shown by solid circles, and with CaZt concentrations buffered with EGTA or EDTA shown by open circles. Each point is the mean of measurements on 2 to 4 sperm samples. +
vated in solutions containing 0.5 mM MgATP and 1 or 2 mM Mg2+. This figure presents data from older experiments with 4 sperm samples, using EGTA or EDTA to buffer Ca2 concentrations, and newer experiments with 3 sperm samples, using BAPTA to buffer the intermediate range of Ca2 concentrations. Figure 3 summarizes the effect of Ca2+ concentration on the asymmetry of flagellar bending waves of demembranated spermatozoa of S. purpurutus reactivated in solutions containing 0.1 mM MgATP and 0.3 mM Mg2+, using BAPTA to buffer the intermediate range of Ca2 concentrations. This figure contains data from measurements on three sperm samples, with two preparations from each sperm sample at each Ca2+ concentration. Each sample yielded a qualitatively similar response curve, with a rise below pCa 7 and a plateau above pCa 7 before a larger rise starting at pCa 6. Similar results were obtained in earlier experiments using only EGTA and EDTA to buffer Ca2+ concentrations (not shown). The calmodulin antagonist, trifluoperazine (TFP), at concentrations up to 50 FM, did not inhibit the asymmetry response to high Ca2+ concentrations. At pCa 4.75,O. 1 mM MgATP, and 0.3 mM Mg2’, 50 FM TFP increased asymmetry from 2.02 2 0.37 rad (n = 35) to 2.53 2 0.55 rad (n = 29). This increase was associated with a 20-25% decrease in frequency and sliding velocity, which appeared to be a photodynamic effect of TFP and prevented examination of the effects of higher TFP concentrations on asymmetry. An affinity-purified antibody against bovine testes calmodulin (Biomedical Tech-
9
8
7
6
5
4
PCa Fig. 3. Mean values of flagellar bending wave asymmetry for experiments with 3 different samples of S. purpurutus spermatozoa, reactivated with 0.1 mM MgATP and 0.3 mM M g z + . Results with CaZt concentrations buffered with EGTA or EDTA are shown by open circles, and with C a 2 + concentrations buffered with BAPTA are shown by solid circles. Each point represents measurements on 40 to 49 spermatozoa, from two separate preparations of demembranated spermatozoa. Since one sperm sample had consistently lower values of asymmetry than the other two, normalization was performed by adding 0.2 radians to the asymmetry values from the low sample and subtracting 0.1 radians from the asymmetry values from the other two samples. Standard deviations for the distribution of values of asymmetry in each sample ranged from 0.18 to 0.49 radians.
+
+
+
nologies, Inc., Stoughton, MA 02072), at concentrations up to 20 Fg/ml, also failed to cause any decrease in asymmetry of flagellar bending waves at pCa 4.75 (not shown). Effects of Exogenous Calmodulin on Flagellar Asymmetry
In four additional experiments, summarized in Figure 4, using S. purpurutus spermatozoa at 0.1 mM MgATP under the same conditions as Figure 3, measurements were also made with 1 pg/ml of bovine brain calmodulin added to the reactivation solutions. The difference between the values of asymmetry in the presence of and in the absence of added calmodulin, shown by the solid line at the bottom of Figure 4, shows the effect of Ca2+ concentration on the asymmetry increase caused by the added calmodulin. The previous work with Lytechinus spermatozoa [Brokaw and Nagayama, 19851 demonstrated that the response to exogenous calmodulin decreased as the time of extraction to remove calmodulin was increased, but did not indicate whether this decrease was a decrease in affinity for calmodulin, or a decrease in maximal response. The experiments summarized in Fig. 5 were designed to address that question. Measurements were made of the effect of exogenous calmodulin, in reactivation solutions at pCa 9, following different calmodulin
Calcium Sensors in Flagella
2-
1
0
1-
#’ ,4
p-&;;+y 4-4’
I
I
I
127
I
I
Fig. 4. Results from 4 additional experiments with the same conditions as in Figure 3 are shown by the solid circles, along with measurements from the same 4 experiments in the presence of 1 pg/ml of exogenous calmodulin (open circles). Approximately 20 spermatozoa were measured in each experiment, at each Ca2+ concentration. Means and standard errors for the 4 experiments are shown. The solid line at the bottom is the difference between the upper two curves, and shows the effect of the added calmodulin, as a function of CaZf concentration.
extraction treatments. Potentially asymmetric spermatozoa, prepared without the second extraction with millimolar Ca2+ and Triton X-100, show essentially no response to the addition of exogenous calmodulin to the reactivation solutions. Following 10 sec extraction with millimolar Ca2+ and Triton X-100, the calmodulin response curve shows a half-maximal response at 1.0 pg/ml of calmodulin, while after 60 sec extraction with millimolar Ca2+ and Triton X-100, a calmodulin concentration of 7.3 pg/ml is required for half-maximal response. The maximum asymmetry observed at high concentrations of calmodulin does not decrease, but instead shows a small increase. The sensitivity of these S . purpuratus sperm preparations to calmodulin is less than that observed in the earlier work with Lytechinus spermatozoa, where spermatozoa extracted with high calcium for 30 sec showed a half-maximal response to calmodulin at a calmodulin concentration of about 0.5 to 1 pgiml. Additional experiments provide data suggesting that one should be cautious about interpreting the response of sperm flagellar asymmetry to addition of exogenous calmodulin (Fig. 5 ) in terms of a simple, reversible calmodulin binding equilibrium. In these experiments, spermatozoa were demembranated and extracted with high calcium just as in Figure 5 . After 60 sec of high calcium extraction, calmodulin was added to give a calmodulin concentration of 20 pgiml, and the mixture was incubated for an additional 30 sec before dilution
0.0 1
0
0.1 1 10 100 Calmodulin concentration (pgmlrnl)
Fig. 5. Asymmetry of demembranated sperm flagella reactivated with 0.1 mM MgATP, 0.3 mM Mg2+ at pCa 9, as a function of calmodulin concentration in the reactivation solution. Each point is a measurement of 20 spermatozoa from one experimental preparation. Results from two experiments, using sperm samples from two different sea urchins, have been combined without normalization. Measurements on sperm preparations that were not extracted with millimolar Ca2+ to remove calmodulin are shown by the x’s and solid line. Open circles were obtained after 10 sec extraction to remove calmodulin. Solid circles were obtained after 60 sec extraction to remove calmodulin. The dashed curves for the two sets of data obtained after calmodulin extraction were obtained by non-linear least squares fitting of the equation: asymmetry = c l + c2/(1 + K/[CaM]), as in Brokaw and Nagayama [ 19851. In this equation, cl represents the baseline asymmetry in the absence of added calmodulin, c2 represents the maximum additional asymmetry obtainable by adding calmodulin, and K is equal to the concentration of calmodulin [CaM] that gives a half-maximal increase in asymmetry.
into reactivation solution. The dilution into reactivation solution without added calmodulin reduces the concentration of exogenous calmodulin to 0.2 pg/ml. However, as shown in Figure 6, the effect on flagellar bending wave asymmetry of the brief incubation with 20 pg/ml calmodulin is much greater than the effect of addition of 0.2 pg/ml calmodulin to the reactivation solutions.
DISCUSSION Two Ca2+ Sensors in Sea Urchin Sperm Flagella
Measurements at 0.5 pCa intervals, summarized in Figures 2 and 3, have identified a new feature of the response of flagellar asymmetry to Ca2 ion concentration that was not detected in earlier work. This feature is the plateau of asymmetry in the vicinity of lop7 M C a 2 + , separating a rise of asymmetry with Ca2+ concentration in the vicinity of M Ca2’ from another rise in asymmetry at Ca2+ ion concentrations greater than lo-‘ M. In experiments using EDTA and EGTA for Ca2+ buffering, this plateau is seen in the lower range for Ca2+ buffering by EDTA, where uncertainty about contaminating Ca2 concentrations, for instance, would +
+
Brokaw
128
3.0 r
0.0
’
gradual response over 4 decades of Ca2 ion concentration. The response of the high-affinity Ca2+ sensor is the same, regardless of whether the sperm flagella have been extracted with high-Ca2+ solutions to remove calmodulin [Brokaw, 19791 and is not altered by the addition of exogenous calmodulin (Fig. 4),which simply adds a constant asymmetry increment to the response at calcium concentrations below lop7 M Ca2+. There is therefore no reason to think that calmodulin is in any way involved in this high-affinity response. This conclusion differs from that of Brokaw and Nagayama [ 19851. Under the conditions of their experiments with Lytechinus spermatozoa, only the high-affinity response could be observed, but it was calculated that this response could be obtained with a calmodulin-like Ca2 sensor under conditions where only 1% of the calmodulin was converted to its active conformation by binding Ca2+.These differences in interpretation are inevitable when saturation of the asymmetry response can appear at Ca2 concentrations below those that saturate Ca2+ binding by a Ca2 sensor. This saturation of the measured asymmetry response can result from saturation of the asymmetry capability of the flagella or from inability to measure the bending waves accurately when the asymmetry becomes extreme. The asymmetry response at Ca2 concentrations greater than M is consistent with the possibility that an endogenous calmodulin-like calcium sensor is responsible for this response. A portion of the flagellar calmodulin remains tightly bound to the axoneme after the extraction with millimolar Ca2’ and Triton X-100 that is required to produce ‘‘potentially symmetric” flagella [Brokaw and Nagayama, 19851. Exogenous calmodulin might interact with the same protein that is the target for the action of this endogenous calmodulin-like calcium sensor. As suggested previously [Brokaw and Nagayama, 19851, a small fraction of the exogenous calmodulin will be in the active conformation, even at very low Ca2+ concentrations, and a small increase in this fraction, beginning at about lop6 M Ca”, may be sufficient to give the largest measurable asymmetry response at Ca2+ concentrations below those that are needed to saturate Ca2+ binding by calmodulin. Additional experiments that attempted to examine the possibility that the low-affinity calcium sensor is related to calmodulin, by using two probes that might be expected to interfere with a calmodulin-mediated process, provided no useful information. Therefore, there is no evidence that the low-affinity sensor is calmodulin, except for the observations that the response is in the proper range of Ca2+ concentrations and that these axonemes contain calmodulin [Brokaw and Nagayama, 19851. +
I
I
I
I
0.1
1
10
100
Calmodulin concentration (pgm/ml) Fig. 6. Effect of calmodulin addition during Ca-extraction on calmodulin response of demembranated sperm flagella, measured as in Figure 5 . The solid line has been transferred from Figure 5 , for the response to calmodulin after 60 sec Ca-extraction. The open circles summarize additional measurements after 60 sec Ca-extraction, showing mean values obtained with the two sperm samples used for these experiments. The solid circles, with standard error bars, were obtained by adding 20 pgiml calmodulin to the Ca-extraction mixture after 60 sec, and then incubating for an additional 30 sec, before the 1:lOO dilution into reactivation solution. Each of these points is the mean of values obtained from 3 or 4 preparations, with approximately 20 spermatozoa measured in each preparation.
be maximal. Solutions buffered with EGTA show the rise in asymmetry near lo-* M Ca2+, but cannot be used at high enough Ca2+ concentrations to show that this rise levels off in the region around lop7 M Ca2+. The plateau detected in experiments with EDTA and EGTA was therefore somewhat questionable. This uncertainty has been removed by experiments using BAPTA to buffer Ca2+ ion concentrations. At the pH and Mg2+ ion concentrations used for these experiments, the optimal Ca2 buffering range for BAPTA is in the vicinity of 10p6.5M Ca2+. Measurements in this range using reactivation solutions buffered with BAPTA confirm the presence of a plateau in this region, and also show a drop in asymmetry at the lower end of this region that merges with the results obtained with EGTA-buffered solutions at lower Ca2 ion concentrations. These results therefore suggest that the change in the asymmetry of flagellar bending waves over a range of 4 decades of Ca2+ ion concentration [Brokaw et al., 19741 is the result of two separate Ca2+ responses. There appears to be a high-affinity Ca2+ sensor that is responsible for changes in asymmetry in the range of lo-’ to M C a Z + ,and a lower-affinity Ca2+ sensor that is responsible for changes in asymmetry at Ca2 concentrations above lop6M. Even if the affinities of these two postulated sensors were closer together, so that there was no detectable plateau between their response ranges, multiple sensors would still be required to explain the +
+
+
+
+
+
+
Calcium Sensors in Flagella
The Response to Exogenous Calmodulin
The results in Figure 5 provide a more detailed view of the decrease in responsiveness of the flagella to calmodulin with increased time of incubation in the presence of millimolar Ca2+ and Triton X-100. As noted previously, with L. pictus spermatozoa, both the degree of asymmetry measured at low Ca2+ in the absence of exogenous calmodulin and the response to exogenous calmodulin decrease with this incubation time and are dependent on the presence of calmodulin-binding agents such as Triton X- 100 or trifluoperazine during incubation [Okuno and Brokaw, 1980; Brokaw and Nagayama, 19851. A reasonable conclusion is that all of the effects of incubation with millimolar Ca2+ and Triton, including release of calmodulin from the flagella [Brokaw and Nagayama, 19851, result from a progressive decrease in the calmodulin binding affinity of the flagellar axonemes. This conclusion does not explain why treatment to reduce the calmodulin binding affinity is necessary to convert the demembranated sea urchin sperm axonemes to a state in which they produce bending waves with values of asymmetry as low as those seen in live, undemembranated, sperm flageIla. One possibility [Brokaw and Nagayama, 19851 is that demembranation releases endogenous calmodulin from a state in which it does not have access to the target protein of the lower-affinity Ca2+ sensor. This calmodulin may be present for some function, such as Ca2 concentration buffering, different from the function of the lower-affinity Ca2+ sensor discussed above. An alternative explanation would be that demembranation causes a sudden increase in calmodulin affinity of some axonemal component, which is then reversed by incubation with high Ca2+ concentrations. These speculations, cast in terms of calmodulin affinity of the axoneme, would be more appropriate if there were evidence for a simple, reversible equilibrium binding of calmodulin during these experiments. However, the results in Figure 6 indicate that the situation may be more complicated. In the presence of Triton X-100, exposure to calmodulin, either low concentrations in the presence of EGTA [Okuno and Brokaw, 1980; Brokaw and Nagayama, 19851 or high concentrations in the presence of Ca2+ (as in the experiment in Fig. 6), can to some extent reverse the effects of extraction with millimolar Ca2 and Triton X- 100. It is not clear whether this reversal involves an irreversible rebinding of calmodulin, or some other change in the axoneme that determines its subsequent sensitivity to exogenous calmodulin. +
+
Comparisons With Other Cilia and Flagella
The presence of calmodulin-like calcium binding proteins and a modulation of ATP-reactivated motility of
129
demembranated cell models or axonemes by changes in Ca2 concentration appear to be common features of cilia and flagella [reviewed by Otter, 19891. In most of the cases studied, a change in motility behavior occurs over a range of about 2 decades of Ca2+ concentration, in the vicinity of M Ca2+. Well known examples are the reversal of swimming direction of cell models of Paramecium and Tetrahymena [Naitoh and Kaneko, 1972; Izumi and Miki-Noumura, 19851; the arrest of lateral cilia of lamellibranch gills [Tsuchiya, 1977; Walter and Satir, 19781; and the flagellar reversal response of Chlamydomonas [Hyams and Borisy, 1978; Bessen et al., 19801. In such cases, a single calmodulin-like calcium sensor could explain the Ca2' concentration dependence of the response. Attempts to support this idea by experiments with calmodulin antagonists have been only partially successful. Reed et al. [ 19821 obtained a partial reversal of calcium-induced arrest of gill cilia using 25 pM TFP. With Chlamydomonas axonemes, Witman and Minervi [1982] found no effect of calmodulin antagonists on flagellar waveform, although high concentrations reduced the number of motile axonemes, an effect that is difficult to relate specifically to calmodulin. The most extensive studies have examined the effects of TFP and other calmodulin antagonists on the ciliary reversal of Triton-extracted cell models of Paramecium that is obtained at calcium concentrations in the range of pCa 5 to 6. With some preparations of Paramecium models, TFP in the 25 to 100 pM range was effective in reversing this effect of calcium, but with other preparations there was no effect of TFP [Otter et al., 1984; Nakaoka et al., 1984; Izumi and Nakaoka, 19871. Izumi and Nakaoka [ 19871 found that CAMP, at relatively high concentrations, could reverse the effect of calcium, or, at lower concentrations, sensitize the cilia to reversal of the calcium effect by TFP. They also observed that when TFP was effective in overcoming ciliary reversal by calcium, the effect of TFP could be inhibited by an inhibitor of CAMP-dependent protein kinase, and they suggested that the effect of calcium and calmodulin in this system might involve activation of a protein phosphatase [Izumi and Nakaoka, 19871. These reports indicate that experiments on complex systems with calmodulin antagonists such as TFP are not simple and definitive. In some cases, effects of much lower Ca2+ concentrations on flagellar motility have also been reported. With trout spermatozoa, Okuno and Morisawa [ 19891 found that motility could be reactivated well at pCa values of 8.5 or higher, but was inhibited when the calcium concentration was greater than pCa 8. Nakaoka et al. [1984] observed changes in swimming velocity and ciliary beat frequency of Paramecium cell models in the range of pCa 6.5 to 7.5, below the range of concentrations where ciliary reversal occurred (pCa 5.5 to 6.5). +
130
Brokaw
With demembranated cell models of Chlamydomonus, Kamiya and Witman [ 19841 observed a transition in the relative activities of the cis and trans flagella in the range between pCa 9 and pCa 7. In similar studies on the trans flagellum expressed by the uni-1 mutant of Chlumydomonas, Omoto and Brokaw [ 19851 determined that changes in frequency and asymmetry did not occur in the same range of Ca2 concentrations. These authors also observed that the bending patterns of flagella on live cells more closely resembled those obtained with demembranated models near pCa 6 rather than near pCa 9, similar to the situation seen here with S . purpuratus spermatozoa (Table I). These examples suggest that the presence of one or more high-affinity calcium sensors in flagella is not unique to sea urchin sperm flagella, but may be a general property of flagella and cilia. So far, sea urchin sperm flagella are somewhat unique in that the same parameter, bending wave asymmetry, appears to be modulated both by low- and high-affinity calcium sensors. +
ACKNOWLEDGMENTS
I thank Holly Dodson, Larry Jones, Sandra Nagayama, and James Pacheco for valuable assistance in the laboratory. This work has been supported by NIH grant GM 18711.
REFERENCES Bessen, M., Fay, R.B., and Witman, G.B. (1980):Calcium control of waveform in isolated flagellar axonemes of Chlamydomonas. J. Cell Biol. 86:446-455. Brokaw, C.J. (1979): Calcium-induced asymmetrical beating of Triton-demembranated sea urchin sperm flagella. J. Cell Biol. 82:401-41 I . Brokaw, C.J. (1984): Automated methods for estimation of sperm flagellar bending parameters. Cell Motil. 4:417-430. Brokaw, C.J. (1986): Sperm Motility. In Schroeder, T.E. (ed.): “Echinoderm Gametes and Embryos” (Methods in Cell Biology, Vol. 27). New York: Academic Press, pp. 41-56. Brokaw, C.J. (1990): Computerized analysis of flagellar motility by digitization and fitting of film images with straight segments of equal length. Cell Motil. Cytoskeleton 17:309-316. Brokaw,, C.J., Josslin, R., and Bobrow, L. (1974): Calcium ion regulation of flagellar beat symmetry in reactivated sea urchin spermatozoa. Biochem. Biophys. Res. Commun. 58:795 - 800. Brokaw, C.J., and Nagayama, S . (1985): Modulation of the asymmetry of sea urchin sperm flagellar bending by calmodulin. J. Cell Biol. 100: 1875-1883. Gibbons, B.H., and Gibbons, I.R. (1980): Calcium-induced quiescence in reactivated sea urchin sperm. J . Cell Biol. 84:13-27. Harrison, S.M., and Bers, D.M. (1987): The effect of temperature and ionic strength on the apparent Ca-affinity of EGTA and the
analogous Ca-chelators BAPTA and dibromo-BAPTA. Biochim. Biophys. Acta 925:133-143. Hyams, J.S., and Borisy, G.G. (1978): Isolated flagellar apparatus of Chlamydomonas: characterization of forward swimming and alteration of waveform and reversal of motion by calcium ions in vifro. J. Cell Sci. 33:235-253. Izumi, A,, and Miki-Noumura, T. (1985): Tetrahymena cell model exhibiting Ca-dependent behavior. Cell Motil. 5:323-33 1. Izumi, A,, and Nakaoka, Y. (1987): CAMP-mediated inhibitory effect of calmodulin antagonists on ciliary reversal of Paramecium. Cell Motil. Cytoskeleton 7:154-159. Kamiya, R., and Witman, G.B. (1984): Submicromolar levels of calcium control the balance of beating between the two flagella in demembranated models of Chlumydomonas. J. Cell Biol. 98: 97- 107. Naitoh, Y., and Kaneko, H. (1972): Reactivated triton extracted models of Paramecium: Modification of ciliary movement by calcium ions. Science 176523,524. Nakaoka, Y., Tanaka, H., and Oosawa, F. (1984): Ca2+-dependent regulation of beat frequency of cilia in Paramecium. J. Cell Sci. 65:223-231. Okuno, M., and Brokaw, C.J. (1980): Effects of Triton-extraction conditions on beat symmetry of sea urchin flagella. Cell Motil. 11363-370. Okuno, M., and Morisawa, M. (1989): Effects of calcium on motility of rainbow trout sperm flagella demembranated with Triton X-100. Cell Motil. Cytoskeleton 14:194-200. Omoto, C.K., and Brokaw, C.J. (1985): Bending patterns of Chlamydomonas flagella: 11. Calcium effects on reactivated Chlamydomonas flagella. Cell Motil 553-60. Omoto, C.K., and Brokaw, C.J. (1989): 2-chloro adenosine triphosphatc as substratc for sca urchin axoncmal movement. Cell Motil. Cytoskeleton 13:239-244. Otter, T. (1989): Calmodulin and the control of flagellar movement. In Warner, F.D., Satir, P., and Gibbons, I.R. (eds.): “Cell Movement, Volume 1: The Dynein ATPases.” New York: Alan R. Liss, Inc., pp. 281-298. Otter, T., Satir, B.H., and Satir, P. (1984): Trifluoperazine-induced changes in swimming behavior of Paramecium: Evidence for two sites of drug action. Cell Motil. 4:249-267. Reed, W., Lebduska, S . , and Satir, P. (1982): Effects of trifluoperazine upon the calcium-dependent ciliary arrest response of freshwater mussel gill lateral cells. Cell Motil. 2:405-427. Tombes, R.M., Brokaw, C.J., and Shapiro, B.M. (1987): Creatine kinase-dependent energy transport in sea urchin spermatozoa. Biophys. J. 52:75-86. Tsien, R.Y. (1980): New calcium indicators and buffers with high selectivity against magnesium and protons: design, synthesis, and properties of prototype structures. Biochemistry 19:23962404. Tsuchiya, T. (1977): Effects of calcium ions on triton-extracted lamellibranch gill cilia: ciliary arrest response in a model system. Comp. Biochem. Physiol [A] 56:353-361. Walter, M.F., and Satir, P. (1978): Calcium control of ciliary arrest in mussel gill cells. J. Cell Biol. 79:llO-120. Ward, G.E., Brokaw, C.J., Garbers, D.L., and Vacquier, V.D. (1985): Chemotaxis of Arbacia puncrulura spermatozoa to resact, a peptide from the egg jelly layer. J. Cell Biol. 101:23242329. Witman, G.E., and Minervi, N . (1982): Role of calmodulin in the flagellar axoneme: Effect of phenothiazines on reactivated axonemes of Chlamvdomonas. Cell Motil. [Suppl.] I : 199-204.
Calcium Sensors in Sea Urchin Sperm Flagella Charles J. Brokaw Division of Biology, California Institute of Technology, Pasadena
The asymmetry of ATP-reactivated flagellar bending waves of Triton-demembrated sea urchin spermatozoa has been measured over a range of free Ca2+ ion concentrations from to lop4 M. Detailed examination of the gradual response of asymmetry to Ca2+ ion concentration over this wide range indicates the presence of two Ca2+ sensors. A high-affinity sensor operates at Ca2+ concentrations near M. A lower-affinity sensor operates at Ca2+ concentrations above M, in the typical range for calmodulin-mediated responses. Incubation of demembranated sperm flagella at high Ca2+ concentrations to release calmodulin is required to enable these Ca2+ responses to be observed. This treatment also causes a decrease in the apparent affinity of the flagella for calmodulin, as determined by measuring the increase in asymmetry in response to addition of exogenous calmodulin at low Ca2+ Concentration. Key words: calmodulin, motility, spermatozoa
INTRODUCTION The ATP-reactivated movement of Triton-demembranated sea urchin sperm flagella can be regulated by the Ca2+ ion concentration of the reactivation solutions, with higher Ca2+ ion concentrations causing increased asymmetry of the flagellar bending waves, without any other major changes in bending wave parameters [Brokaw et al., 1974; Brokaw, 19791. Although a function for this Ca2+ sensitivity has not been definitively established, its involvement in chemotactic response of spermatozoa to egg jelly peptides has been suggested [Ward et al., 19851. The response to Ca2+ ion concentrations is difficult to study with demembranated sea urchin sperm preparations that are prepared exclusively using solutions containing EGTA to maintain very low Ca2+ ion concentrations (“potentially asymmetric” spermatozoa [Gibbons and Gibbons, 19801). Even when reactivated at very low Ca2+ concentrations, such sperm preparations display a high level of bending wave asymmetry, which increases to levels that are too extreme to measure accurately as the Ca2+ concentration is increased [Brokaw, 19791. Exposure to millimolar Ca2+ ion concentrations during or after demembranation can produce “potentially symmetric” sperm flagella that generate nearly 0 1991 Wiley-Liss, Inc.
symmetric flagellar bending waves at low Ca2+ ion concentrations and demonstrate gradually increasing asymmetry as the Ca2+ ion concentrations in the reactivation solutions are increased [Brokaw et al., 19741. This effect of millimolar Ca2 concentrations is time and temperature dependent, requires the presence of detergent [Okuno and Brokaw, 19801, results in release of calmodulin from the sperm flagella, and sensitizes the flagella to exogenous calmodulin [Brokaw and Nagayama, 19851. The exploration of the calmodulin and calcium sensitivity of flagella by Brokaw and Nagayama [ 19851 utilized spermatozoa of Lytechinus pictus. Spermatozoa from this sea urchin species show, in addition to increased asymmetry, an attenuation or damping of bending wave amplitude in the distal region of the flagellum as Ca2+ ion concentration is increased [Okuno and Brokaw, 19801. This effect interferes with measurement of bending wave asymmetry at higher Ca2+ concentrations. Spermatozoa from another sea urchin, Strongylocentrotus purpurutus, show much less increase in bend+
Received July 20, 1990; accepted October 25, 1990. Address reprint requests to Dr. C.J. Brokaw, Division of Biology 156-29, Caltech, Pasadena, CA 91 125.
124
Brokaw
ing wave attenuation at high Ca2+ ion concentrations. However, initial attempts to demonstrate calmodulinsensitivity of S. purpurutus spermatozoa following extraction with millimolar Ca2 solutions indicated that neither the procedures used for L . pictus spermatozoa [Brokaw and Nagayama, 19851 nor the procedures used earlier for demembranation and reactivation of S. purpurutus spermatozoa [e.g., Brokaw, 19861 gave good results. Development of a more optimal procedure for preparation of “potentially symmetric” sperm flagella with good calmodulin sensitivity from S. purpurutus has now made possible a more detailed examination of the relationship between flagellar asymmetry and concentrations of Ca2+ ions and calmodulin in this species. In particular, the ability of these flagella to respond gradually to changes in Ca2+ ion concentration over a range of more than 4 decades of Ca2+ ion concentration [Brokaw et al., 19741 has been examined more fully, since this behavior does not resemble that expected for a simple Ca-binding receptor. This behavior has been clarified by making measurements at intervals of 0.5 pCa unit, rather than just at intervals of 1 pCa unit, and by using a new calcium buffer, BAPTA, to confirm results obtained with solutions buffered with EGTA and EDTA. This increased data collection has been made possible by the evolution of efficient computerized methods for analysis of photographic records of sperm flagellar bending. +
MATERIALS AND METHODS
The methods used here for preparation and reactivation of spermatozoa from the sea urchin, S. purpurutus, differ from earlier methods used in this laboratory [Brokaw, 19861 but have been partially described in work reported by Tombes et al. [1987] and by Omoto and Brokaw [1989]. The sea urchins were induced to shed gametes by injection of 0.6 M KC1, and “dry” spermatozoa were collected from the aboral surface of the urchin, diluted with an equal volume of cold 0.5 M NaCl, and stored on ice until used. Demembranation was performed by mixing 1 p1 of the stock sperm suspension with 50 pl of demembranation solution containing 0.15 M KC1, 10 mM Tris buffer, 1 mM dithiothreitol (DTT), 2 mM MgSO,, 0.2 mM EGTA, 0.2 mM ATP, 10 pM CAMP, and 0.04% Triton X-100. This solution, and all of the other solutions used in this work, were adjusted to pH 8.2. In contrast to spermatozoa of L. pictus, spermatozoa from S. purpuru~ K consistently S show good reactivation without exposure to solutions containing CAMP, and it has not been possible to show any consistent benefits of including CAMP and ATP in the demembranation solutions unless the pH of the demembranation solution is reduced to about 7.0. These components were included routinely in the de-
membranation solution with the hope that this would increase the uniformity of the results obtained with different sperm samples. After 30 seconds incubation in demembranation solution, 450 pl of Ca-extraction solution containing 0.15 M KCl, 10 mM Tris buffer, 1 mM DTT, 2.2 mM CaCl,, and 0.01% Triton X-100 was added, and the mixture was incubated for an additional 30 sec. Ten microliters of this extracted sperm suspension was then diluted into 1.0 ml of reactivation solution for observations. All reactivation solutions contained 0.25 M potassium acetate, 10 or 20 mM Tris buffer, 1 mM DTT, and 0.5% polyethyleneglycol. ATP, MgSO,, CaCl,, EGTA, EDTA, and BAPTA were added as needed for particular experiments. Most of the experiments reported here used reactivation solutions containing 0.1 mM MgATP2+ and 0.3 mM Mg2+. Recipes for these solutions, containing desired Ca2+ ion concentrations, were calculated as described by Brokaw [1986], using the constants given there for EDTA and EGTA. The constants given by Tsien [1980] for BAPTA were used or modified for the higher ionic strength of the reactivation solutions used here, leading to a value of 6.5 for the pK,, for BAPTA [Harrison and Bers, 19871. These calculations assumed a level of calcium contamination in the solutions of 20 pM. EGTA, EDTA, and bovine brain calmodulin were purchased from Sigma and BAPTA was from Calbiochem. A drop of demembranated sperm suspension in reactivation solution was placed on a microscope slide and illuminated with darkfield strobe illumination, using a temperature-controlled stage at 18°C in a room maintained at 18°C. Photographs were taken at a magnification of 80 X and a flash rate of 120 or 150 flashes per second, on film moving at 0.25 m/sec in a 35 mm oscilloscope camera [Brokaw, 19861. A series of images from each photograph was digitized, either manually as in earlier work or with a new computerized digitization procedure [Brokaw, 19901, to obtain data sets that were analyzed by computerized methods [Brokaw, 19841 to obtain parameters for the flagellar bending waves, as in earlier work. RESULTS Parameters of Flagellar Motility of S. purpuratus Spermatozoa
Table I presents unpublished data from the experiments of Tombes et al. [ 19871 on live spermatozoa of S. purpurutus. Seven sperm samples were measured under control conditions, in artificial sea water at 18°C. This table also presents data for seven samples of demembranated spermatozoa, reactivated at an MgATP concentration of 0.5 mM and intermediate Ca2 concentration +
Calcium Sensors in Flagella
125
TABLE I. Parameters of Sperm Flagellar Motility Demembranated and 0.5 mM MgATP-reactivated spermatozoa (7 samples; 417 cells)
Live spermatozoa (7 samples; 153 cells) Mean of sample means
Std. error of mean
Mean sample std. dev.
Mean of samDle means
Std. error of mean
Mean sample std. dev.
38.4 1.84 180 28.7 0.35 1.18
0.5 0.05 6 1.3 0.03 0.06
2.7 0.11 15 2.0 0.17 0.21
33.6 2.15 185 28.1 0.27 1.33
0.3 0.06 5 0.6 0.05 0.11
1.7 0.19 16 1.8 0.19 0.28
Frequency (s-I) Bend angle (rad) Sliding velocity (rad s - ' ) ~ Wavelength (pm) Attenuation' Asymmetry (rad)'
"Velocity of metachronal sliding in propagating bends; average magnitude for principal and reverse bends. 'Attenuation measures decrease in curvature of bends as they propagate along flagellum [Tombes et al., 19871. 'Asymmetry calculated as difference between absolute values of principal and reverse bend angles [Brokaw, 19791.
(pCa 6.5). Under these conditions of demembranation and reactivation, the parameters of reactivated motility are very similar to those of intact spermatozoa, except for a somewhat lower frequency, which is combined with an increased bend angle so that there is no significant difference in sliding velocity. In the previous study [Brokaw, 19791, data were presented showing that Ca2 concentration had very little effect on parameters of flagellar motility, other than the asymmetry of the bending waves. Those experiments were performed at 0.1 mM MgATP, at relatively low beat frequencies (ca. 10 Hz) because the reactivation solutions contained methyl cellulose and KCl rather than K acetate. In the experiments reported here, at 0.1 mM MgATP with 0.3 mM Mg2+, the beat frequencies were approximately 19 Hz. As shown in Figure 1, under these conditions also, there is very little variation in sliding velocity, wavelength, or bending wave attenuation over the range from pCa 9.25 to 4.75. There is a slight decrease (ca. 10%) in beat frequency at higher Ca2+ concentrations (not shown), but this is compensated by an increase in bend angle, so that the sliding velocity remains essentially constant. However, at the higher MgATP concentration of 0.5 mM, which gives beat frequencies and sliding velocities more comparable to those of live spermatozoa, there are significant changes in some of these other bending wave parameters with pCa, especially at pCa values less than 7 (Fig. 1). To avoid complications resulting from these changes, effects of Ca2+ concentration on asymmetry were evaluated at 0.1 mM MgATP as well as at 0.5 mM MgATP.
0
8
+
The Effect of Ca2+ Ion Concentration on Flagellar Asymmetry
Figure summarizes the effect Of tion on the asymmetry of flagellar bending of demembranated spermatozoa of S . purpuratus reacti-
I 20 I
5 .c1
2 5 2
0 0
I
I
I
I
I
r
0
0.8 0.6
0
:
0
0.4
0
c.
-
0.2 0.0
1 I
250 r
0 Q)
a*
200-
2 2 150-
v
_0 0
5
>
100-
I
I
I
I
1
I
9
8
7
6
5
4
PCa Fig. 1. Effects of Ca'+ concentration on some bending wave parameters of ATP-reactivated spermatozoa from S. purpuratus. Solid points are from experiments at 0.1 mM MgATP; open points are from experiments at 0.5 mM MgATP. Each point represents a mean for measurements on approximately 40 spermatozoa. Values of bending wave asymmetry from these experiments are included in Figures 2 and 3.
126
Brokaw 0 0
S 0
O
0 0
0 0
0 ”
01 9
8
5
6
7
0
0
4
PCa
I
10
Fig. 2. Mean values of flagellar bending wave asymmetry from experiments with 7 different samples of S. purpurutus spermatozoa, reactivated with 0.5 mM MgATP and 1 or 2 mM M g 2 + . Results from experiments with 4 sperm samples, with Ca2 concentrations buffered with EGTA at pCa 7 and higher, and with EDTA at pCa less than 7 are shown by open squares. Results from experiments with another 3 sperm samples are shown by circles, with Ca2+ concentrations buffered with BAPTA shown by solid circles, and with CaZt concentrations buffered with EGTA or EDTA shown by open circles. Each point is the mean of measurements on 2 to 4 sperm samples. +
vated in solutions containing 0.5 mM MgATP and 1 or 2 mM Mg2+. This figure presents data from older experiments with 4 sperm samples, using EGTA or EDTA to buffer Ca2 concentrations, and newer experiments with 3 sperm samples, using BAPTA to buffer the intermediate range of Ca2 concentrations. Figure 3 summarizes the effect of Ca2+ concentration on the asymmetry of flagellar bending waves of demembranated spermatozoa of S. purpurutus reactivated in solutions containing 0.1 mM MgATP and 0.3 mM Mg2+, using BAPTA to buffer the intermediate range of Ca2 concentrations. This figure contains data from measurements on three sperm samples, with two preparations from each sperm sample at each Ca2+ concentration. Each sample yielded a qualitatively similar response curve, with a rise below pCa 7 and a plateau above pCa 7 before a larger rise starting at pCa 6. Similar results were obtained in earlier experiments using only EGTA and EDTA to buffer Ca2+ concentrations (not shown). The calmodulin antagonist, trifluoperazine (TFP), at concentrations up to 50 FM, did not inhibit the asymmetry response to high Ca2+ concentrations. At pCa 4.75,O. 1 mM MgATP, and 0.3 mM Mg2’, 50 FM TFP increased asymmetry from 2.02 2 0.37 rad (n = 35) to 2.53 2 0.55 rad (n = 29). This increase was associated with a 20-25% decrease in frequency and sliding velocity, which appeared to be a photodynamic effect of TFP and prevented examination of the effects of higher TFP concentrations on asymmetry. An affinity-purified antibody against bovine testes calmodulin (Biomedical Tech-
9
8
7
6
5
4
PCa Fig. 3. Mean values of flagellar bending wave asymmetry for experiments with 3 different samples of S. purpurutus spermatozoa, reactivated with 0.1 mM MgATP and 0.3 mM M g z + . Results with CaZt concentrations buffered with EGTA or EDTA are shown by open circles, and with C a 2 + concentrations buffered with BAPTA are shown by solid circles. Each point represents measurements on 40 to 49 spermatozoa, from two separate preparations of demembranated spermatozoa. Since one sperm sample had consistently lower values of asymmetry than the other two, normalization was performed by adding 0.2 radians to the asymmetry values from the low sample and subtracting 0.1 radians from the asymmetry values from the other two samples. Standard deviations for the distribution of values of asymmetry in each sample ranged from 0.18 to 0.49 radians.
+
+
+
nologies, Inc., Stoughton, MA 02072), at concentrations up to 20 Fg/ml, also failed to cause any decrease in asymmetry of flagellar bending waves at pCa 4.75 (not shown). Effects of Exogenous Calmodulin on Flagellar Asymmetry
In four additional experiments, summarized in Figure 4, using S. purpurutus spermatozoa at 0.1 mM MgATP under the same conditions as Figure 3, measurements were also made with 1 pg/ml of bovine brain calmodulin added to the reactivation solutions. The difference between the values of asymmetry in the presence of and in the absence of added calmodulin, shown by the solid line at the bottom of Figure 4, shows the effect of Ca2+ concentration on the asymmetry increase caused by the added calmodulin. The previous work with Lytechinus spermatozoa [Brokaw and Nagayama, 19851 demonstrated that the response to exogenous calmodulin decreased as the time of extraction to remove calmodulin was increased, but did not indicate whether this decrease was a decrease in affinity for calmodulin, or a decrease in maximal response. The experiments summarized in Fig. 5 were designed to address that question. Measurements were made of the effect of exogenous calmodulin, in reactivation solutions at pCa 9, following different calmodulin
Calcium Sensors in Flagella
2-
1
0
1-
#’ ,4
p-&;;+y 4-4’
I
I
I
127
I
I
Fig. 4. Results from 4 additional experiments with the same conditions as in Figure 3 are shown by the solid circles, along with measurements from the same 4 experiments in the presence of 1 pg/ml of exogenous calmodulin (open circles). Approximately 20 spermatozoa were measured in each experiment, at each Ca2+ concentration. Means and standard errors for the 4 experiments are shown. The solid line at the bottom is the difference between the upper two curves, and shows the effect of the added calmodulin, as a function of CaZf concentration.
extraction treatments. Potentially asymmetric spermatozoa, prepared without the second extraction with millimolar Ca2+ and Triton X-100, show essentially no response to the addition of exogenous calmodulin to the reactivation solutions. Following 10 sec extraction with millimolar Ca2+ and Triton X-100, the calmodulin response curve shows a half-maximal response at 1.0 pg/ml of calmodulin, while after 60 sec extraction with millimolar Ca2+ and Triton X-100, a calmodulin concentration of 7.3 pg/ml is required for half-maximal response. The maximum asymmetry observed at high concentrations of calmodulin does not decrease, but instead shows a small increase. The sensitivity of these S . purpuratus sperm preparations to calmodulin is less than that observed in the earlier work with Lytechinus spermatozoa, where spermatozoa extracted with high calcium for 30 sec showed a half-maximal response to calmodulin at a calmodulin concentration of about 0.5 to 1 pgiml. Additional experiments provide data suggesting that one should be cautious about interpreting the response of sperm flagellar asymmetry to addition of exogenous calmodulin (Fig. 5 ) in terms of a simple, reversible calmodulin binding equilibrium. In these experiments, spermatozoa were demembranated and extracted with high calcium just as in Figure 5 . After 60 sec of high calcium extraction, calmodulin was added to give a calmodulin concentration of 20 pgiml, and the mixture was incubated for an additional 30 sec before dilution
0.0 1
0
0.1 1 10 100 Calmodulin concentration (pgmlrnl)
Fig. 5. Asymmetry of demembranated sperm flagella reactivated with 0.1 mM MgATP, 0.3 mM Mg2+ at pCa 9, as a function of calmodulin concentration in the reactivation solution. Each point is a measurement of 20 spermatozoa from one experimental preparation. Results from two experiments, using sperm samples from two different sea urchins, have been combined without normalization. Measurements on sperm preparations that were not extracted with millimolar Ca2+ to remove calmodulin are shown by the x’s and solid line. Open circles were obtained after 10 sec extraction to remove calmodulin. Solid circles were obtained after 60 sec extraction to remove calmodulin. The dashed curves for the two sets of data obtained after calmodulin extraction were obtained by non-linear least squares fitting of the equation: asymmetry = c l + c2/(1 + K/[CaM]), as in Brokaw and Nagayama [ 19851. In this equation, cl represents the baseline asymmetry in the absence of added calmodulin, c2 represents the maximum additional asymmetry obtainable by adding calmodulin, and K is equal to the concentration of calmodulin [CaM] that gives a half-maximal increase in asymmetry.
into reactivation solution. The dilution into reactivation solution without added calmodulin reduces the concentration of exogenous calmodulin to 0.2 pg/ml. However, as shown in Figure 6, the effect on flagellar bending wave asymmetry of the brief incubation with 20 pg/ml calmodulin is much greater than the effect of addition of 0.2 pg/ml calmodulin to the reactivation solutions.
DISCUSSION Two Ca2+ Sensors in Sea Urchin Sperm Flagella
Measurements at 0.5 pCa intervals, summarized in Figures 2 and 3, have identified a new feature of the response of flagellar asymmetry to Ca2 ion concentration that was not detected in earlier work. This feature is the plateau of asymmetry in the vicinity of lop7 M C a 2 + , separating a rise of asymmetry with Ca2+ concentration in the vicinity of M Ca2’ from another rise in asymmetry at Ca2+ ion concentrations greater than lo-‘ M. In experiments using EDTA and EGTA for Ca2+ buffering, this plateau is seen in the lower range for Ca2+ buffering by EDTA, where uncertainty about contaminating Ca2 concentrations, for instance, would +
+
Brokaw
128
3.0 r
0.0
’
gradual response over 4 decades of Ca2 ion concentration. The response of the high-affinity Ca2+ sensor is the same, regardless of whether the sperm flagella have been extracted with high-Ca2+ solutions to remove calmodulin [Brokaw, 19791 and is not altered by the addition of exogenous calmodulin (Fig. 4),which simply adds a constant asymmetry increment to the response at calcium concentrations below lop7 M Ca2+. There is therefore no reason to think that calmodulin is in any way involved in this high-affinity response. This conclusion differs from that of Brokaw and Nagayama [ 19851. Under the conditions of their experiments with Lytechinus spermatozoa, only the high-affinity response could be observed, but it was calculated that this response could be obtained with a calmodulin-like Ca2 sensor under conditions where only 1% of the calmodulin was converted to its active conformation by binding Ca2+.These differences in interpretation are inevitable when saturation of the asymmetry response can appear at Ca2 concentrations below those that saturate Ca2+ binding by a Ca2 sensor. This saturation of the measured asymmetry response can result from saturation of the asymmetry capability of the flagella or from inability to measure the bending waves accurately when the asymmetry becomes extreme. The asymmetry response at Ca2 concentrations greater than M is consistent with the possibility that an endogenous calmodulin-like calcium sensor is responsible for this response. A portion of the flagellar calmodulin remains tightly bound to the axoneme after the extraction with millimolar Ca2’ and Triton X-100 that is required to produce ‘‘potentially symmetric” flagella [Brokaw and Nagayama, 19851. Exogenous calmodulin might interact with the same protein that is the target for the action of this endogenous calmodulin-like calcium sensor. As suggested previously [Brokaw and Nagayama, 19851, a small fraction of the exogenous calmodulin will be in the active conformation, even at very low Ca2+ concentrations, and a small increase in this fraction, beginning at about lop6 M Ca”, may be sufficient to give the largest measurable asymmetry response at Ca2+ concentrations below those that are needed to saturate Ca2+ binding by calmodulin. Additional experiments that attempted to examine the possibility that the low-affinity calcium sensor is related to calmodulin, by using two probes that might be expected to interfere with a calmodulin-mediated process, provided no useful information. Therefore, there is no evidence that the low-affinity sensor is calmodulin, except for the observations that the response is in the proper range of Ca2+ concentrations and that these axonemes contain calmodulin [Brokaw and Nagayama, 19851. +
I
I
I
I
0.1
1
10
100
Calmodulin concentration (pgm/ml) Fig. 6. Effect of calmodulin addition during Ca-extraction on calmodulin response of demembranated sperm flagella, measured as in Figure 5 . The solid line has been transferred from Figure 5 , for the response to calmodulin after 60 sec Ca-extraction. The open circles summarize additional measurements after 60 sec Ca-extraction, showing mean values obtained with the two sperm samples used for these experiments. The solid circles, with standard error bars, were obtained by adding 20 pgiml calmodulin to the Ca-extraction mixture after 60 sec, and then incubating for an additional 30 sec, before the 1:lOO dilution into reactivation solution. Each of these points is the mean of values obtained from 3 or 4 preparations, with approximately 20 spermatozoa measured in each preparation.
be maximal. Solutions buffered with EGTA show the rise in asymmetry near lo-* M Ca2+, but cannot be used at high enough Ca2+ concentrations to show that this rise levels off in the region around lop7 M Ca2+. The plateau detected in experiments with EDTA and EGTA was therefore somewhat questionable. This uncertainty has been removed by experiments using BAPTA to buffer Ca2+ ion concentrations. At the pH and Mg2+ ion concentrations used for these experiments, the optimal Ca2 buffering range for BAPTA is in the vicinity of 10p6.5M Ca2+. Measurements in this range using reactivation solutions buffered with BAPTA confirm the presence of a plateau in this region, and also show a drop in asymmetry at the lower end of this region that merges with the results obtained with EGTA-buffered solutions at lower Ca2 ion concentrations. These results therefore suggest that the change in the asymmetry of flagellar bending waves over a range of 4 decades of Ca2+ ion concentration [Brokaw et al., 19741 is the result of two separate Ca2+ responses. There appears to be a high-affinity Ca2+ sensor that is responsible for changes in asymmetry in the range of lo-’ to M C a Z + ,and a lower-affinity Ca2+ sensor that is responsible for changes in asymmetry at Ca2 concentrations above lop6M. Even if the affinities of these two postulated sensors were closer together, so that there was no detectable plateau between their response ranges, multiple sensors would still be required to explain the +
+
+
+
+
+
+
Calcium Sensors in Flagella
The Response to Exogenous Calmodulin
The results in Figure 5 provide a more detailed view of the decrease in responsiveness of the flagella to calmodulin with increased time of incubation in the presence of millimolar Ca2+ and Triton X-100. As noted previously, with L. pictus spermatozoa, both the degree of asymmetry measured at low Ca2+ in the absence of exogenous calmodulin and the response to exogenous calmodulin decrease with this incubation time and are dependent on the presence of calmodulin-binding agents such as Triton X- 100 or trifluoperazine during incubation [Okuno and Brokaw, 1980; Brokaw and Nagayama, 19851. A reasonable conclusion is that all of the effects of incubation with millimolar Ca2+ and Triton, including release of calmodulin from the flagella [Brokaw and Nagayama, 19851, result from a progressive decrease in the calmodulin binding affinity of the flagellar axonemes. This conclusion does not explain why treatment to reduce the calmodulin binding affinity is necessary to convert the demembranated sea urchin sperm axonemes to a state in which they produce bending waves with values of asymmetry as low as those seen in live, undemembranated, sperm flageIla. One possibility [Brokaw and Nagayama, 19851 is that demembranation releases endogenous calmodulin from a state in which it does not have access to the target protein of the lower-affinity Ca2+ sensor. This calmodulin may be present for some function, such as Ca2 concentration buffering, different from the function of the lower-affinity Ca2+ sensor discussed above. An alternative explanation would be that demembranation causes a sudden increase in calmodulin affinity of some axonemal component, which is then reversed by incubation with high Ca2+ concentrations. These speculations, cast in terms of calmodulin affinity of the axoneme, would be more appropriate if there were evidence for a simple, reversible equilibrium binding of calmodulin during these experiments. However, the results in Figure 6 indicate that the situation may be more complicated. In the presence of Triton X-100, exposure to calmodulin, either low concentrations in the presence of EGTA [Okuno and Brokaw, 1980; Brokaw and Nagayama, 19851 or high concentrations in the presence of Ca2+ (as in the experiment in Fig. 6), can to some extent reverse the effects of extraction with millimolar Ca2 and Triton X- 100. It is not clear whether this reversal involves an irreversible rebinding of calmodulin, or some other change in the axoneme that determines its subsequent sensitivity to exogenous calmodulin. +
+
Comparisons With Other Cilia and Flagella
The presence of calmodulin-like calcium binding proteins and a modulation of ATP-reactivated motility of
129
demembranated cell models or axonemes by changes in Ca2 concentration appear to be common features of cilia and flagella [reviewed by Otter, 19891. In most of the cases studied, a change in motility behavior occurs over a range of about 2 decades of Ca2+ concentration, in the vicinity of M Ca2+. Well known examples are the reversal of swimming direction of cell models of Paramecium and Tetrahymena [Naitoh and Kaneko, 1972; Izumi and Miki-Noumura, 19851; the arrest of lateral cilia of lamellibranch gills [Tsuchiya, 1977; Walter and Satir, 19781; and the flagellar reversal response of Chlamydomonas [Hyams and Borisy, 1978; Bessen et al., 19801. In such cases, a single calmodulin-like calcium sensor could explain the Ca2' concentration dependence of the response. Attempts to support this idea by experiments with calmodulin antagonists have been only partially successful. Reed et al. [ 19821 obtained a partial reversal of calcium-induced arrest of gill cilia using 25 pM TFP. With Chlamydomonas axonemes, Witman and Minervi [1982] found no effect of calmodulin antagonists on flagellar waveform, although high concentrations reduced the number of motile axonemes, an effect that is difficult to relate specifically to calmodulin. The most extensive studies have examined the effects of TFP and other calmodulin antagonists on the ciliary reversal of Triton-extracted cell models of Paramecium that is obtained at calcium concentrations in the range of pCa 5 to 6. With some preparations of Paramecium models, TFP in the 25 to 100 pM range was effective in reversing this effect of calcium, but with other preparations there was no effect of TFP [Otter et al., 1984; Nakaoka et al., 1984; Izumi and Nakaoka, 19871. Izumi and Nakaoka [ 19871 found that CAMP, at relatively high concentrations, could reverse the effect of calcium, or, at lower concentrations, sensitize the cilia to reversal of the calcium effect by TFP. They also observed that when TFP was effective in overcoming ciliary reversal by calcium, the effect of TFP could be inhibited by an inhibitor of CAMP-dependent protein kinase, and they suggested that the effect of calcium and calmodulin in this system might involve activation of a protein phosphatase [Izumi and Nakaoka, 19871. These reports indicate that experiments on complex systems with calmodulin antagonists such as TFP are not simple and definitive. In some cases, effects of much lower Ca2+ concentrations on flagellar motility have also been reported. With trout spermatozoa, Okuno and Morisawa [ 19891 found that motility could be reactivated well at pCa values of 8.5 or higher, but was inhibited when the calcium concentration was greater than pCa 8. Nakaoka et al. [1984] observed changes in swimming velocity and ciliary beat frequency of Paramecium cell models in the range of pCa 6.5 to 7.5, below the range of concentrations where ciliary reversal occurred (pCa 5.5 to 6.5). +
130
Brokaw
With demembranated cell models of Chlamydomonus, Kamiya and Witman [ 19841 observed a transition in the relative activities of the cis and trans flagella in the range between pCa 9 and pCa 7. In similar studies on the trans flagellum expressed by the uni-1 mutant of Chlumydomonas, Omoto and Brokaw [ 19851 determined that changes in frequency and asymmetry did not occur in the same range of Ca2 concentrations. These authors also observed that the bending patterns of flagella on live cells more closely resembled those obtained with demembranated models near pCa 6 rather than near pCa 9, similar to the situation seen here with S . purpuratus spermatozoa (Table I). These examples suggest that the presence of one or more high-affinity calcium sensors in flagella is not unique to sea urchin sperm flagella, but may be a general property of flagella and cilia. So far, sea urchin sperm flagella are somewhat unique in that the same parameter, bending wave asymmetry, appears to be modulated both by low- and high-affinity calcium sensors. +
ACKNOWLEDGMENTS
I thank Holly Dodson, Larry Jones, Sandra Nagayama, and James Pacheco for valuable assistance in the laboratory. This work has been supported by NIH grant GM 18711.
REFERENCES Bessen, M., Fay, R.B., and Witman, G.B. (1980):Calcium control of waveform in isolated flagellar axonemes of Chlamydomonas. J. Cell Biol. 86:446-455. Brokaw, C.J. (1979): Calcium-induced asymmetrical beating of Triton-demembranated sea urchin sperm flagella. J. Cell Biol. 82:401-41 I . Brokaw, C.J. (1984): Automated methods for estimation of sperm flagellar bending parameters. Cell Motil. 4:417-430. Brokaw, C.J. (1986): Sperm Motility. In Schroeder, T.E. (ed.): “Echinoderm Gametes and Embryos” (Methods in Cell Biology, Vol. 27). New York: Academic Press, pp. 41-56. Brokaw, C.J. (1990): Computerized analysis of flagellar motility by digitization and fitting of film images with straight segments of equal length. Cell Motil. Cytoskeleton 17:309-316. Brokaw,, C.J., Josslin, R., and Bobrow, L. (1974): Calcium ion regulation of flagellar beat symmetry in reactivated sea urchin spermatozoa. Biochem. Biophys. Res. Commun. 58:795 - 800. Brokaw, C.J., and Nagayama, S . (1985): Modulation of the asymmetry of sea urchin sperm flagellar bending by calmodulin. J. Cell Biol. 100: 1875-1883. Gibbons, B.H., and Gibbons, I.R. (1980): Calcium-induced quiescence in reactivated sea urchin sperm. J . Cell Biol. 84:13-27. Harrison, S.M., and Bers, D.M. (1987): The effect of temperature and ionic strength on the apparent Ca-affinity of EGTA and the
analogous Ca-chelators BAPTA and dibromo-BAPTA. Biochim. Biophys. Acta 925:133-143. Hyams, J.S., and Borisy, G.G. (1978): Isolated flagellar apparatus of Chlamydomonas: characterization of forward swimming and alteration of waveform and reversal of motion by calcium ions in vifro. J. Cell Sci. 33:235-253. Izumi, A,, and Miki-Noumura, T. (1985): Tetrahymena cell model exhibiting Ca-dependent behavior. Cell Motil. 5:323-33 1. Izumi, A,, and Nakaoka, Y. (1987): CAMP-mediated inhibitory effect of calmodulin antagonists on ciliary reversal of Paramecium. Cell Motil. Cytoskeleton 7:154-159. Kamiya, R., and Witman, G.B. (1984): Submicromolar levels of calcium control the balance of beating between the two flagella in demembranated models of Chlumydomonas. J. Cell Biol. 98: 97- 107. Naitoh, Y., and Kaneko, H. (1972): Reactivated triton extracted models of Paramecium: Modification of ciliary movement by calcium ions. Science 176523,524. Nakaoka, Y., Tanaka, H., and Oosawa, F. (1984): Ca2+-dependent regulation of beat frequency of cilia in Paramecium. J. Cell Sci. 65:223-231. Okuno, M., and Brokaw, C.J. (1980): Effects of Triton-extraction conditions on beat symmetry of sea urchin flagella. Cell Motil. 11363-370. Okuno, M., and Morisawa, M. (1989): Effects of calcium on motility of rainbow trout sperm flagella demembranated with Triton X-100. Cell Motil. Cytoskeleton 14:194-200. Omoto, C.K., and Brokaw, C.J. (1985): Bending patterns of Chlamydomonas flagella: 11. Calcium effects on reactivated Chlamydomonas flagella. Cell Motil 553-60. Omoto, C.K., and Brokaw, C.J. (1989): 2-chloro adenosine triphosphatc as substratc for sca urchin axoncmal movement. Cell Motil. Cytoskeleton 13:239-244. Otter, T. (1989): Calmodulin and the control of flagellar movement. In Warner, F.D., Satir, P., and Gibbons, I.R. (eds.): “Cell Movement, Volume 1: The Dynein ATPases.” New York: Alan R. Liss, Inc., pp. 281-298. Otter, T., Satir, B.H., and Satir, P. (1984): Trifluoperazine-induced changes in swimming behavior of Paramecium: Evidence for two sites of drug action. Cell Motil. 4:249-267. Reed, W., Lebduska, S . , and Satir, P. (1982): Effects of trifluoperazine upon the calcium-dependent ciliary arrest response of freshwater mussel gill lateral cells. Cell Motil. 2:405-427. Tombes, R.M., Brokaw, C.J., and Shapiro, B.M. (1987): Creatine kinase-dependent energy transport in sea urchin spermatozoa. Biophys. J. 52:75-86. Tsien, R.Y. (1980): New calcium indicators and buffers with high selectivity against magnesium and protons: design, synthesis, and properties of prototype structures. Biochemistry 19:23962404. Tsuchiya, T. (1977): Effects of calcium ions on triton-extracted lamellibranch gill cilia: ciliary arrest response in a model system. Comp. Biochem. Physiol [A] 56:353-361. Walter, M.F., and Satir, P. (1978): Calcium control of ciliary arrest in mussel gill cells. J. Cell Biol. 79:llO-120. Ward, G.E., Brokaw, C.J., Garbers, D.L., and Vacquier, V.D. (1985): Chemotaxis of Arbacia puncrulura spermatozoa to resact, a peptide from the egg jelly layer. J. Cell Biol. 101:23242329. Witman, G.E., and Minervi, N . (1982): Role of calmodulin in the flagellar axoneme: Effect of phenothiazines on reactivated axonemes of Chlamvdomonas. Cell Motil. [Suppl.] I : 199-204.
